Projects
![]()
No projects found
Investors
New
DAO Treasuries
DATs
Lend / Borrow APRs
Chains
Categories
Spaces
Agentic Commerce, On-Chain Prediction Markets, and the Future of Autonomous Trading
Recorded: March 16, 2026 Duration: 0:57:55Space Recording
Full Transcription
Thank you. all right we're live edison thanks for joining us today how are you doing hey i'm doing fine
yourself yeah really good thank you i've been excited for this one it feels like we've got
about four or
five different reports in the works now covering basically every touch point that you guys cover so
i've been excited to get into this one even from a selfish perspective just to pick your brain on
a lot of stuff but uh yeah i think i think what would be worth doing is just setting the scene
of how you guys have got to this point currently because there's a million different ways we can take this one from
agentic payments agentic commerce hip for prediction markets the future perhaps in sport
what's happening with like traditional equities commodities on chain like there's a million
different ways i want to get into this but i think it's worth setting the scene with regards
to like how you guys have came online what the initial starting idea was and how you got to this point now okay um i think that i could kind of like just start off with what base is and uh what we are
building next for the for the next few few months or years to come so um this is a trading venue that allows people to basically trade everything and spend everywhere.
So at the core, we believe that by building a very strong self-custodial wallet,
people could assess any markets that they want,
whether this is about crypto, equities, commodities, and prediction market.
equities, commodities and prediction market.
We had this idea since last year when I think that
people don't really think about equities to be on chain.
There's a lot of attention and
and excitement to bring equities on chain back in last year.
But I think that before the rise of the SYZ market, there hasn't been one that is so strong
that what we are seeing right now.
I think that with the current Iranian situation, we see that people are actually trading a
lot of oil.
And I think that is really something that was really hard to predict in last year.
But we kind of had this idea that markets and trading should be permissionless.
And people in all over the world, they should be able to get it regardless of their nationality, regardless of where they're going from.
So, and of course, I mean, I've been in crypto for many years, like I think close crypto user experience has to really improve a lot.
And that is also one of the motivations as to why we built BASE.
So BASE is about trying to really make the best experience that's possible around trading.
Like if you think about it as a crypto wallet, there's many crypto wallets out there.
There's probably hundreds or thousands of wallets.
Any software that holds private key is a crypto wallet.
But why base is so strong is because we really optimize our whole experience to be around
traders.
We basically use the best APIs to give very reliable data feeds.
And I think that's really what makes it to be what it is today.
So a trading venue is not just about being able to assess markets, but it's also about
trying to do like payments.
Like for example, if you were to use like interactive brokers
or like Robinhood,
on-ramping is one part of the experience.
If the on-ramp is slow, it's clunky or it's expensive,
it basically adversely impact your user experience.
So that was also why we,
I think that we have done quite a lot on that front,
basically made trading to be a lot more accessible for users.
Awesome.
Yeah, so you touched on the traditional markets,
in particular oil,
given the backdrop of what is happening
in the Middle East at the minute of what is happening in the Middle
East at the minute. And you're starting to see the traditional, even the traditional media outlets
reporting on what's happening on top of hyperliquid and particularly the oil markets, which is kind of
insane to think of even six to ten months ago that that would even be a thing. It was all like the idea was there, but the execution and just kind of like the weird
external events that have allowed that to kind of be front and center and not to kind of,
I'd rather the events weren't happening, but like a kind of like culmination of like a lot of preparation
and a lot of like forward thinking markets bringing that online and now everyone is using hyperliquid and these on-chain markets is a de facto way of checking
what's actually happening over weekends so like how much activity has been on your platform with
regards to this are you seeing interest and volumes come in what's the open interest starting
to look like when did this start to tick up was there like a bit of a oh god there's something
worth paying attention to here with regards to equities commodities indices precious metals and
things like that because it feels like the top assets being traded on chain at the minute are
definitely not on chain native assets um i think that uh i mean I don't message me or like
they don't appear to be people who will be using based I mean because these
people you use like centralized exchanges a lot and they started to reach out to
me and say that oh like what the difference like in the different markets
because like on hyperliquid is there's a permissionless market. So there are a few HIP3 providers
that provide markets for the same asset.
And you're asking like, what is the kind of differences?
And I think that's when like,
I feel that there's like an onboarding opportunity
where people are trying to trade commodities
that is otherwise not available on the venues
that they are usually trading.
I also see the same thing for people who don't usually come onto crypto but because they
want to trade silver.
I think that it's probably the same kind of users that are coming to trade oil and given that hyperliquid is like one of the
platforms that allows people to trade oil over the weekend you can trade between cr or like brand
oil i think that gives people a lot of like opportunity and space to really speculate on
the development situation at iran. And do you think like,
I don't know how you guys think about this,
but the talk,
like your messaging to attract new users
towards those markets.
Do you,
do you have to give different messaging
to more of a crypto native audience
than you have to give to maybe people
that you just,
just as you say,
interested in trading silver or trading oil?
Is it like,
is it a different targeted approach with regards to marketing when you do that
i think this is an interesting question um i think in fact i believe that this is an area that
we should be doing more of we should be doing things that are better at onboarding like users
so what i've seen is that uh of these users, they have not really
interacted much with Hyperliquid. I think it's the same kind of situation that we
had since the beginning, so we are not really new to it. When people were using
based, they see based to be the exchange. They don't really see like the back and behind uh in a way i think that this is very
similar to how users basically associate robin hood to be uh when something goes wrong on robin
hood you don't say that okay like it must be the infrastructure behind robin hood that is
that is down but you look to a robin hood and say that like i try to place a trade but this
trade does not work um i i think it's the same situation in base as well that when people come
to us they see us as the venue even though we basically build on top like hyperliquid so um
so there were some confusions that i think that we should be looking to address over time
uh for example there's there are, there are many conflicting tickles
when it comes to oil. You have US oil, you have oil, you have CL, you have Brent oil,
and sometimes it is being paired with different code assets. Sometimes it's USDC, USDT, and also USDH.
Sometimes transferring between one asset to the other has a little bit of slippage. So I think those are things that we should be working on as well as like
making it as cheap as possible for people to deposit money from the bank.
I think interactive brokers, for example, they do zero fees.
I think like interactive brokers, for example, they do like zero fees.
I think in our version, it's very hard to do at zero fees, but if we cut down the fees
to all the way to, let's say, 5 bps or 10 bps, I think it's going to make a big difference
between how, for example, if people would do on-ramp from using MoonPay, it's going to
cost 5% to six percent
and that's not really good looking to fly and do you have to think about how you know you say about
like different markets obviously hyper liquid and different settlements it's like i'd assume this
would be my theory not knowing like behind the scenes with you guys but i think for a mobile
first crowd you're probably gonna have to lean more
into like asset abstraction and trying to like really dumb down that ui ux so if they do want
to just express interest in oil or in the s p or in gold then that the complexities of the markets
behind that and the settlement currencies behind that in a mobile first environment would probably
lend itself well to just being completely abstracted away.
But then I'd imagine, I'm totally guessing here,
but your desktop users are going to definitely be more,
right, I'll manually sift through which markets I actually want to trade,
what I want to settle in.
You're used to having a little bit more friction
and can kind of put up with that.
But I don't know if there's any different kind of user behaviors
across the app
as opposed to across what you see on the users on the desktop.
I think that you touched on a very interesting question as to how much
should we really abstract away from users.
I think in a way, I think that base got really far because we put the eye,
I mean, we put what pro users users are looking for and pro users want
a lot of control over what they want to trade on.
So I think that's kind of different from the retail users.
Retail users, you can tell them that this is the market, this is only one market.
If you want to trade oil, there's only one market that you can trade on.
So I think that we have some internal discussions around that.
Have not had anything finalized.
I mean, we would love to have something finalized on like how we should be,
whether we should be having an opinion on what markets people should be trading on. For the time being, Bayes is a permissionless market.
You could search across anything that is on Hyperliquid as long as it's on the strict
list that Hyperliquid has.
So Hyperliquid maintains a strict list and I think that the requirements to be there
is to make sure that the liquidity, the volume is being there.
So that list is being maintained by hyperliquid, we simply tap into it.
So we have not had a very full-fledged discussion on how strongly should we have an opinion
on what markets people should be trading. And because we are quite neutral, we don't run our own markets on the commodes.
So there is really nothing in it for me to kind of like direct people to one market over the other.
There's nothing that's commercial that is in our interest to do so.
I think that you are frozen.
Are you all right? I'm going to go ahead and get some more of the stuff I've been doing. Sorry, I don't know what's going on with these stream platforms at the minute
yeah so um the reason the reason i was asking across mobile and desktop is because of last
week's um did you see the 50 million dollar swap on that route through ave and apparently when they traced
it back it was um someone using the mobile app and i wouldn't imagine anyone's performing a 50
million dollar swap on an application that is mobile first um which is a little bit confusing
but um yeah that's that's the reason i asked that one yeah yeah it is, I mean, let me just try, I'm just trying to pull up the latest stats on
our mobile versus like our desktop app.
I think generally people do like, it may sound, I mean, like people normally use BASE because it's a continuous experience.
They use a desktop or laptop to do whatever trade they want.
They can step out, they can go to the gym, they can go for lunch or dinner and still get
access to the position.
If something goes on wrongly, they could always close the position while they're outside.
I think this has been a very strong use case in base.
And yeah, I think that we don't really see
kind of like a difference between what people
should be seeing on web and desktop, web and mobile.
I mean, definitely you should have more features on web,
but the overall
guidance as to what we should be putting it should be similar like for example if we do not
have any opinion on what users should be trading then we should not discriminate that between
mobile and the web i mean it's our stance. is what everyone's trying to wrap their head around. Everyone's releasing a CLI, everyone's looking at how they onboard
maybe the next 100 million users is actually agents
as opposed to what we originally thought
might have just been regular humans.
How are you guys thinking about that?
Because I know you guys have been really pushing.
I've seen a lot of really interesting developments,
of really interesting developments,
but I'd love to know how you're driving that forward with the team.
but I'd love to know how you're driving that forward
with the team.
I would say that when it comes to AI agents,
our team has been on a very,
we have been observing that for a while.
Like it's not that we don't want to innovate,
but we prefer to see,
I think that like,
okay, if we were to kind of like fast uh like backtrack a few
months back last year uh firstly a lot of our attention and work has always been on my wall
and the web experience and that's that has been something that we took quite a while to polish up
uh but we are so focused on that we just really frankly does not have enough time
we just really frankly does not have enough time to be distracted with any other things.
The base team is a very lean team. We are around the same size as a hyperliquid team.
So we don't have a luxury of putting a team of 10 or 20 to figure out the AI strategy for base.
We don't have that luxury to do so. So I think that when
we started to look into it, I think that last year, we don't really see too much. When we
see what people are doing as an agentic wallet, we don't really see that this is something
that is on our agenda. Firstly, I think that back in those days,
a lot of the agentic framework was around Lanchain.
So the idea, Lanchain is basically a framework
where people would say that,
I would love to swap $100 of BTC across like four chains.
And you will figure out like,
they need to divide $ dollars into four and use a
different bridge into the four chains that people want um i don't really see too much assignment in
that i think that this is great for user experience but i mean it really sounds cool on the hackathon
idea but i don't really see that it's going to be used in production by people. And there's a lot of excitement around like people doing CLI tools or MCP.
And we see what those are, but I think
that there should be something that's a lot more powerful
that BASE should be doing.
I think that this strategy has got to be, at least in our opinion,
what we see is that, firstly, OpenClaw has been
like a game changer for some of us in the team. For me, I've been using it for around
like three months right now. And I think that this is like a tool that has reached a level
of maturity that more people should be able to use it. It does a perfect job in memorizing your context across a few places.
And that's really basically like backing to our AI strategy that we feel that an agenting
wallet shouldn't just be being able to place a trade on hyperliquid or polymarket.
These are things that are out there.
You could just do it.
So there's really nothing much that we should be putting our effort there.
In fact, I'll say that the winner is going to be the one that is going to dominate the context of what people are trying to do.
So by context, I mean that it crawls the best information
that's out there on the internet.
It gets it to be a lot faster than usual.
It does not hallucinate.
And it understands the preference
as to what people want to trade,
or the preference around their trading behavior.
So I think that long-term,
we really do want to build this like knowledge graph around
the context that people have that's why we have been focusing a lot on the infrastructure behind
ai agents that if people use it then the inference cost is going to increase
then we will hope that we can capture some of this pie for our users as well
yeah side question.
Do you just use the native OpenClaw memory
or do you use like Memzero or any other additional memory?
Because memory is the biggest issue
I think we're running into at the minute.
Yes, I think for me, I just use the native memory.
But I think that is like,
and of course you could also store some of those nodes on like Obsidian. I think that is like, I, and of course, like you could also store some of those notes
on like obsidian.
I think that's really helpful.
Yeah.
Yeah.
Like, otherwise I think that it's just really quite brutal to be, to be out there, like expecting like an open claw instance to remember everything.
And it wouldn't be a best like use of it.
I mean, in the event that your open clawlaw instance would get corrupted or clean yourself up,
then you lose access to all your information and knowledge.
So for me, I use like Obsidian to come by store all of these nodes.
Good job.
So I suppose like a small segue here
into what we're seeing with x402 um agentic payments i've seen ramp
start to come online now so that they're building towards an agentic payments future i've seen the
stripe team go on like tbpn and say that they're building with this in mind and obviously tempo
i think like tempo's uh token standard is calledIP20, where it's a little bit more programmable at the token level,
like with this in mind.
And then obviously Visa and Mastercard are putting out a lot of work around it.
But how are you seeing how this should progress?
Are the big incumbents missing a trick with not primarily thinking about this
with regards to stablecoins?
Do you think stablecoins and X402 is inherently the right move for eugenic payments?
Or do you think the big incumbents like your Visa and MasterCard of the world
will have something to say about that?
I think that we are in a quite unique scenario that we are both doing things on the stablecoin.
We are both like doing things on the stablecoin.
So today, I would say that last week we released our AI inference as a service.
So people could pay with USDC or USTH with S4 to get access to inference models.
That is on there for infinite tips.
I mean, the idea is not new.
People say that this is like one of the ways that AI agents
could get assessed to compute, to deployment,
and to storage as well.
So we have done that.
It is pretty trivial.
I don't see a lot of things that is wrong with 4.2
at the moment.
Some people wish that 4.2 has the ability for people to do things like
holding a charge for like $10 and releasing that like a week or two later similar to credit cards.
So I think that I'll be hoping that 402 actually takes off well because I don't see anything wrong with that today.
With the Visa and MasterCard, the issue is really around that the adoption is going to be slow.
And I know this is like firsthand because before base become based, we were running a payment gateway. We were going after merchants and trying to get merchants to accept stablecoin as a payment.
And we understand the issues that they have when it comes to accepting stablecoins.
The issues could be all the way from the custodial of their funds to how do they get it into
local currency, to how do they report tax in the country that they are in.
So there's a lot of issues when it comes to merchant adoption.
So I see that to be one of the issues
when it comes to accepting stablecoin.
And of course, the shortcut is to borrow the rails
that Visa and MasterCard has
to complete this payment loop for the agents.
I'm pretty sure you guys have seen agent cards on the market.
We've been prototyping something similar for our users,
that they could delegate their card assets to an agent payment
so that that personal assistant could complete the payment for you.
Those are things that we are looking heavily into.
We're working very closely with Visa.
And I think that right now,
there's like,
like this use case,
I would say it is still quite in the early stages
because some of these websites,
they actively block out agents
that are trying to preview credit card payments on the form. So I think that this
will take some time to resolve. And if that goes goes
through, I see I don't see a reason why Asian cards are not
going to take off. So for the time being, we don't have a very
strong opinion there. We just really hope that if either of
these two work,
base will also do fine because we have both.
Yeah, you'll be in a very good position.
So this is an interesting point, what you say about the merchant side.
So whoever can own that distribution,
and I think we've been talking about this a lot internally,
what I think a lot of people who have kind of primarily just lived in the UK
or primarily just lived in the US or particularly out west
is a lot of the payment infrastructure is completely location dependent
and requires like local activation.
And it probably makes sense for a lot of teams to just target one specific geolocation
and then continue to expand i think revolute have done an amazing job because every single merchant form of
payment what that user flow looks like our payout is totally different so in indonesia everyone uses
curious or if you're in portugal they use like mb where it's like the qr code scanner apple again
that merchant checkout process is completely different in location dependent.
And there's also a weird social dynamic effect as well, where it requires this network effect
to actually take hold so that everyone's on the same payment rails and everyone's just
completely used to what that user checkout experience looks like.
That's why I think if you can get it natively online first with Stripe,
that looks pretty good because it's the native checkout of the internet.
So that makes a ton of sense.
But what do you think people need to do to make that easier for the merchant adoption?
Because even the big stablecoin, the big stablecoin providers like Tether and Circle,
they haven't even figured that out yet.
They're not in every
kind of like mom and pop shop around the world where you can just accept that and that just
becomes like a seamless process and that seems like a huge huge huge undergoing but i feel like
whoever can get out in front of that first and win that battle has got such a huge power law and even
just like the use of stickiness and retention and muscle memory factor there where it's you know it's very very difficult to be toppled if whoever comes
out on top.
If I were to be like I think that as a consumer so how would you want to expedite the adoption
from the merchants?
Like speaking as someone who used to run a payment gateway,
I believe that the way is for merchants to use that stablecoin more for payments.
So there are some merchants that they accept stablecoin.
I think that in the past, this could be like a travel platform,
or it could be maybe like a clothing or like a luxury
clothing store like if they accept stablecoin try to use it as much as
possible like normally what goes on in the mind of this merchants is that when
something when a major brand has a like does a USDC or any stablecoin payment
they are looking to test the demand of that service.
Like in the initial phase, it's not going to be, of course, they will be wondering how
many people really want to use it.
So when a major brand does it, and if you believe that this is like, okay, we should
be expediting crypto adoption worldwide worldwide try to use the stablecoin
because this will give them the reason to see that this is a large uh like i mean there's a lot of
like customers that are using usdc or usdt for payments and uh let's try to support it as much
as we can um so i think that once it passes a certain stage,
then there should be a network effect of like,
there's more merchants accept stablecoin,
then there is more use case of holding more stablecoin
because you want to pay for it.
And I think that this will be a very kind of like
self-fulfilling loop that will be out there.
So I would say this is like what consumers should be
doing. For the merchants, I know that it's going to be hard, but I really do think that Stablecoin
is a much more superior payment going forward. I've really tried and I really hated to use swift transfer it is just so slow to transfer anything and uh yeah and i yeah like if you if
you had to transfer money like overseas like for vendors or like just like uh to pay for things
uh abroad then uh yeah you could really see how appealing it is to have a stablecoin. Yeah. So what's happening currently and what needs to happen on, like, let's just say we're
in, I don't know, we're in a coffee shop, you've got your, you've got your crypto card.
So the conversion is happening before the payment is getting settled.
So like the, the app or like the infrastructure and the swap that's actually enabling that
to happen is settling to a local currency and then the local currency is just
getting paid towards the merchant whereas in the future we'd like to get to a point where it's just
a stable coin to stable coin transaction and then that that doesn't inherently um make the merchant
and or user so for the two three percent that these card payment providers actually
um baking into like the cost of the items there is that what's actually happening with that user flow
there um okay so there are quite a few crypto cards out there um i think that we operate on
the different like uh i mean some of us we have a different working model with
like card providers like but I think that I can just like comment on what we
do because that this is something that I am familiar on I am not familiar on what
other people do but I was super familiar what we do so basically I would say that
we go quite bad matter.
That means we don't really abstract ourselves too much
with partners.
Generally, when you look into card payments,
there are people who take a shortcut.
And they go with someone who kind of
packaged this as a service.
They subscribe to the service.
And they don't have to do a lot of things
when it comes to the authorization and so on.
But in exchange, that middle band or the B2B provider,
they take a lot of spread.
The good thing there is that
they can get the market really fast.
But the bad thing is that
they don't keep a lot of profit
margin in themselves. So for us, because we go a little bit lower, so our tech stack is
a lot more different and it can be interesting. So when someone swipes a card in a store,
so let's say that you use a base card. I have one over here. So this
is my base card. So if I use a base card, this is a go card. You can see yourself here.
If you use a base card to pay, then there will be a charge that is being made on the card.
like a charge that is being made on the card.
So basically it comes to us and say that,
okay, like this guy, Grant,
he is trying to spend $10 on coffee.
And we check his balance and we make sure
that he has at least $10 worth of assets in his account.
Assets could be a combination of your USD
or your Solana or your Hype balance.
And we basically basically lock that funds
and we basically try to,
we have to reply back to the merchant
within two seconds
to say that Grant has the money to do so.
A lot of our process are done on the backend
because there is really no good reason
as to why you would want to risk um the like
making it on chain because if you do it on chain there could be a risk that it hasn't reached a
block confirmation and so on so we we does all this lock-in on the off-chain side and once that
is going through there is a very long process when it comes to settlement. I think that this is a lot to do with the legacy process on Visa, where the payment
actually goes through a few countries and a few providers.
You have, it goes through all the way from merchants to the merchant acquirer, to the
Visa, all the way to the acquiring bank, which is the one that issues you the card before
reaching to the card before reaching to
the card provider.
So it's a very long loop.
I think you can search it on Google on how this works.
But I think that today, a lot of the settlement is still being done in fiat.
So we are still doing it through USD. So I mean, today we are purely doing it based on USD.
Previously, we used to do it on a few different currencies.
And, yeah, and all the things are just like happening behind the scenes.
It is not an instant thing.
So, I think that one misconception that people have is that when I use a card,
the payment is actually done, fully executed, and it's done and sent to the merchant. But it is not.
The merchant only received the payment one or two weeks later. There are some, I would say there are
some legacy things as to why is it one or two weeks, but there are also some pretty legit reasons as to why it's taking so long.
For example, if you were to try to buy something online and you would like to get it refunded,
then the fact that the payment hasn't gone through is the reason why you're getting refunded in the first place.
Because then the refund basically gets reverted before the merchant receives
the payment.
But if all the payments are done instantaneously and the money is now on the merchant's wallet,
I think that your refund story and experience is going to look very different to what it
is today.
So, sometimes adding a little bit of speed bump actually provides a boost in the user experience when it comes to shopping online.
Yeah, so I think that's basically how CryptoCard works.
For us, we receive money in digital assets, whether this is USDC, USDT, or like Selena, or Haik.
And we do our own conversion with our own off-prem partners to get it into local
currencies as cheaply as we can.
Great.
Sorry to bore you. This is like a lot of pictures.
No, no, no. I'm super interested by all this stuff. So quick, quick pivot now, because
I know you guys have got another layer with regards to prediction markets
um are you guys going to be integrating hip for day one like when that goes live how are you kind of thinking about that um yeah let's let's start there because i've got numerous
different questions and rabbit holes to go down i think that for like um hip for right
I think that for like HIP fall, right?
Like it is too intense that.
So I think that that's like, I mean, of course, we try to experiment with it.
We try our best to look into like, how do we integrate that?
Yeah, so of course, we'll be looking at doing it on day one.
But the thing about outcome based market is that it has got to do with a lot of the markets
that people want to trade on.
So I feel that it has got to be time with the news that is newsworthy enough that people
really want to trade on those markets.
Just to give you one example,
so last year, I think many people were trying to trade
on the F1 because they were trying to speculate
on who is the winner and so on.
I think that this year is gonna be like a World Cup.
So people are trying to speculate on the teams
that will be winning on the World Cup.
So I think that outcome-based market
has got to do, has got to have a good timing
that once like there is a newsworthy event
that this is when we roll it out.
At least for us because we already have it,
we already looked into how it's been done on Testnet
and we have the experience on doing a very good user experience when it comes to
prediction markets.
We'll be, we'll love to be able to support it on day one.
Awesome.
The one thing you touched on the, on the World Cup and football, which is outside
of this, my, my other life that i love more than
anything but um the problem the problem is currently which i feel like a lot of people
are trying to solve and i've seen i've spoke to a lot of teams in this in this vertical but
usually the user behavior on football is parlors accumulatorsulators, multi-leg bets.
You guys had any thoughts around that?
And I suppose the extension of that is leverage baked in,
so there's asymmetric payouts on binary outcome markets.
I don't know if you've got any thoughts on that,
or it might just be kind of like a throwaway question.
Yeah, I don't have too much opinion on the market structure of that.
I don't really want to comment on things that I don't have a very good understanding of.
I think at least for us, we try to make the best experience of the things that we can control.
How do people see these markets?
What kind of experience do they want?
What kind of like a hybrid and combination of events
that we could do for people, right?
So just to give you one example,
when it comes to like the football matches,
we could actually have a way about doing
the stream directly on the based web app.
And I think this would be a good outcome when it comes to people
who are trying to put it on the on the web app um i think this is like our age our age is about
trying to see that what are the things that people really want to trade on uh rather than trying to
figure out like what is the best market structure to support outcome-based contracts yeah yeah what's um i used to live
in indonesia for a few years and um what's interesting about the premier league streaming
streaming app called video so this doesn't happen in the west but say if i was to log on to video
and there was like a football game on now on the app they have a chat on the
right so as you can imagine it's just it's absolute chaos everyone's talking about the
game talking about the match but i think what you're saying now is like a really yeah yeah
yeah yeah it's that that is like so foreign in in the west that would should never be a thing on sky or you know tnt or wherever but
that is like a that's a golden hack for prediction markets like it just makes so much sense
i'm i'm glad to hear this from someone in the uk that this is like a ui that you like
i i usually thought that um the way the west and the asia has a very different
The West and the Asia has very different stick on what a good user experience should be.
The West, whether it's in Europe or US, you guys prefer very clean UI.
It's almost like the cleaner UI, the better it is.
But the Asians, we look in a clean UI and we think like, this is like absolutely,
I can't do anything.
Like I'm paralyzed.
Because like, we are just very used to being bombarded with like a lot of information on one screen.
Yeah.
Yeah, a good example of that is
if you go on to like Alibaba, Timu, Shopee,
all of those, they are hyper gamified and it feels like a little bit
again speaking from it from someone who was brought up in the uk the gamification like the
bonuses the everything that they do from like a user retention perspective even though it feels
again from the west that feels out of place to be on those type
of apps but it obviously works like the pop-ups like collecting the points and it all bundles
into one but the western markets are just not used to like used to being exposed to that kind
of thing so it feels like a bit jarring at first but then like if you think about it from a product
manager perspective it makes a ton of sense, like in hyper-competitive markets in particular.
Yeah, I think, yeah,
they just have so much games to play with, right?
They'll give you like,
oh, this is like a daily checking for you, right?
You have to do like three or five or seven check-ins
and you'll get a reward.
Yeah, so like, these are like things that is very, these are things that is very Asian.
And if you look into the first page of the app,
it's just so crowded.
I think it's even more crowded if you were just trying
to see it in English.
Because Asian characters have a much shorter character
density.
So for example, if you're trying to get an uber ride
it's like it's a transport transport is like t-r-e-n-s-p-o-r-t so that's like
nine nine letters right so but in mandarin is basically like two characters yeah so
so so when you try to when you were looking looking at the mobile phone, this is also why the Asian
kind of user experience can be so compact because the language density is really small
compared to English, which is a lot more lengthy.
Yeah.
Yeah.
So I think your Uber example is a good jumping off point for, again, another thing that the
West have been very, very slow to adopt is this idea of a super app.
And I think Elon is slowly trying to push Twitter towards that with X payments rolling
out and they've supercharged the chat function now, which is a hell of a lot better.
But again, the ubers of the
world are slowly getting there the we chats are just another level like absolutely everything you
can kind of think of like your i don't know grabs and gojeks and all those kind of like
pandema all those kind of like applications out out in east they are just like an all-in-one
like they are really a super app like a lot of people say that word and throw it around quite loosely.
But you guys are kind of like slowly moving in position
and now you can kind of see it with the card,
the spot, perps, prediction markets.
I don't know if chat could be a function in the future.
I don't know like how you guys are kind of thinking about that.
But it feels like you're in a very, very, very good position
to try and push towards that idea of what people would associate with a super app.
And don't have you got any kind of unique insight on that one?
I think that initially I would like, I mean, we, we really wanted to become like this super
app of like, um, uh, in, uh, that is natively integrated with trading markets.
So our target line is basically trade everything,
spend everywhere.
The spending portion has always traditionally
been about cut payments.
And I think that we basically try to look into
the possibility of doing a e-commerce store ourself.
So we call it Base Mall. I think that when it comes to e-commerce,
I feel that this is something that we may have some age there
given that, like my other co-founder,
he used to work in Alibaba.
So he was actually building an e-commerce app
when he was there.
So we really think that there's some market anger
when it comes to a super app that people could be doing
trading of different sets and having a cut they could spend.
And even potentially, we start to own the distribution
as to what people want to buy.
So that is what we were initially like we were thinking about so we
started work on basically like doing the backend trying to do the fulfillment of items
uh we start off with backend because we think that the e-commerce app of the future may not look like what it is on Alibaba.
We feel that it may actually look a lot like publicity.
That if you have an intent, then we would recommend you things to buy.
So the chat app becomes the central of the experience of the entire store app.
That this is like an app that allows you to assess not just markets, get news.
But if you lack anything, let's say that if you are trying to trade and then you
will say that, okay, what would George Soros do?
And maybe our chat app will just tell you that you should
just go and buy a book from George Soros
and see what he thinks, and you can place
an order on Amazon. So that
could be one of the use case that is out there.
But it is not just that.
Sometimes I think that
as I was using OpenClaw a lot,
I think that initially
I would just be using it
just for strictly crawling the news,
like the things that people say.
But after a while, it became my context engine,
that it knows a lot of things about what I'm thinking about.
And even the things that I should be doing on my to-do list.
For example, let's say that this could be
finding a place for the next team launch. Or this could be buying a new laptop
or buying a new desktop for running a server.
And I think it would be awesome
that we could combine and tie this back
to placing a trade on Taobao or Amazon.
Like thinking that Taobao or Amazon, that could be like, what if we are like the builder
code on top of Amazon?
So I think there was like the thinking there.
Yeah.
Yeah.
I love that.
I've seen a lot of people rolling out skills and plugins basically because Amazon has kind of turned into this huge seo game
right and like all the descriptions and titles are like i think it's called aeo so it's like a
whole subsect of an industry where the titles are hyper optimized like a lot of it is drop shipping
so it's just basically like how how much can you optimize the title and description and and the reviews to kind of
bring your margin down and so a lot of people have been building kind of like what you're saying
there with like the likes of claude and saying i just want um i don't know creatine gummies right
i want creatine gummies but i don't want anything that has been hyper optimized the reviews look a
little bit dodgy but the delivery time has to be
like less than a week and then like the return from that plugin should just be like a list this
is the price it's the product this is like the reviews this is the location so quick you can get
it so you're actually what you would manually do when you're searching amazon to actually get to
the product that you actually wanted it's a way more refined process of actually getting to the
point you want and then final point of what you're saying there is actually enabling the checkout experience of
that as well so i feel like it's all coming and it's you've seen the open ai integration with
shopify and and stuff like that as well so yeah it's all moving a bit too quick for even like me
who has a team to like capture all this stuff and report on this stuff so i don't know how you guys
are keeping up a bit as well we are barely keeping up like like there's like this meme right that's like i wake up there's a
new technological advancement i kind of like feel like i'm a cat like when i wake up there's like a
new thing that's coming up um just very difficult to keep up um but but it is also it is also exciting and tiring at the same time.
You get what I mean?
Because we all know that, at least I'm very pro-AI.
I think that AI is going to take over a lot of jobs.
It may sound trivial right now, but there
are some people who still is anti-AI. I know of companies that they are still banning people
from using AI on coding.
So I wonder if I'm still living in a bubble.
But I think AI is just basically
bring a huge chunk of productivity for people so and i think that um
it's scary to think of it that maybe like if you don't really embrace it that people always
like stress that you will be left behind i think it's like it's kind of a formal that if you don't
if you don't set up your open claw today then you'll be left behind yeah i can i can never really get a good
reading or reference point of like are we just in this twitter echo chamber and like you know
the meme of like the average person on the street just doesn't even know that the markets have been
crashing for the past like six months and they're living in total bliss and i have no i have no idea what like the average person is doing like are they just on the free chat gpt or
gemini plan or whatever they use and like how far down the rabbit hole are some people like i have
i have no idea i'm completely in the dark around that i'd have a guess but i feel like i'd be
completely wrong i think that like um i mean speaking as someone that is like i i mean i really feel that the
people that you meet in crypto are the early adopters of that that technology curve i'm not
sure what what is the curve but there's like this curve where there's like early adopters and then like the
like and then like the mainstream and so on i i feel that many crypto people they are on the early adopters like cycle uh typically the older that you see them uh that if they are like the
class of like 20 30 they're like super og they are the kind of people that will want to go through
all kinds of hoops to try to get something, right?
So I've met the people, I mean, for me personally,
I came into Tuklo in 2018.
I wouldn't consider myself ERR, like to be very early
compared to those that were in 2014, 2015.
Like those people, they have to go through like the
dark internet they have to pay people some money and they get BTC in their
wallet or they have a service like GPU farm in their dom and they try to mine
Bitcoin or Dogecoin and that's why they get their first cryptocurrencies.
And surprisingly right those people are the ones that is currently doing a lot
on ai like i know of someone who is like a very early like dogecoin like minor like even though
today he's not doing too much crypto but he is like heavily trying to optimize on how it's like
agentic like coding and so on i think that we might not really be living in the bubble,
but maybe this might really be the early adoption
as to what is to come.
It is really scary when you see how strong agents can be
and how fast they develop.
At least for me, I try to use Cursor in the early days.
The very, very early days when it was just a chat,
there's no composer. And I recall back in those days, like the very, very early days when it was just a chat, there's no composer.
And, um, I recall back in those days and Reddit was saying that this is like completely not
usable.
The tab is not usable and it gives like shit code and so on.
But you first thought, just like a year and a half later, like the world is very different.
Right.
So people are just like just doing coding across multiple terminals
and sessions.
And I just feel that this adoption
will just accelerate even faster than what it is today.
So yeah, so I think that models are actually
just improving so much.
People are subsidizing models.
And it might just be an infection point like this year
nice so edison is there anything i might have missed that you want to make people aware of
because i appreciate we covered a lot of topics but your product covers a lot of different areas
that are all worth and i podcasting and of them cell phones but if there's anything you want to
make people aware of then the floor is yours um i think that like for for us is that uh
we we really see that uh that the base that we we see today is going to get a lot more users
and it is just sometimes that those users are not really humans you know it's like this side
the users could be agents that agents could be using it to do things that we
don't we can't foresee today so like I think that the call is that like humans
can get a bank account it just could not so if the war has like that let's say
I'm just picking this up like I'll think that if the world has like tribulent like
better accounts in the world,
then there should be a factor
of like at least 100 times or 200 times more
bank accounts in the Asian world.
And that's because every Asian also needs
their own money to kind of like complete some tasks, right?
So if I'm trying to like basically
run training strategies, just one of it.
But if I want to kind of like have an agent that helps me do
collect money for school fees,
if I'm a principal, I want to collect money for school fees,
then this agent could actually just be,
just be one that kind of like message like people
and say that, oh, please pay me here.
And then he gets his own like smaller account
that he could hold onto it.
So it's like, you know, like in school,
you have like this like the kids that is like
the treasury manager and they kind of like go around
collecting like smaller money here and there.
I kind of like see the same thing for like agents as well,
that agents will be the,
is going to be the one that will do things for
that we cannot foresee today so um i feel like we are in this very very interesting technology
standpoint and uh whoever that's going to win may be the people who are going to own the
infrastructure that's that is like powering it so and that's what we want to be absolutely so um we'll have to do this again
maybe next quarter or the quarter after see how far you guys have progressed how many more products
that you guys have got but uh i really enjoyed the past hour a lot of good insight there so
thanks so much for coming on and uh yeah anyone who is watching we'll see you tomorrow with the
lefi team yeah again more agentic stuff seems like it's all encompassing at the minute.
So yeah, see you tomorrow.
Over and out.
Take it easy.
Bye.
Yeah, see you okay.
Bye-bye.
